Bitcoin Forum

What do you guys think about this?

What happens if any GPU hit the market with 2 GH/s with solar powered unit? (Yeah, fantasy! just like cell phones in 80's)

GPU mining won't die.

It'll just go to razor-thin profit margins and won't be worthwhile for many people to do.

Lol it's already at that point, yet some people continue to do so.

Gotta say I'm pretty surprised at the rate of which difficulty is still rapidly increasing at, even when the price was sub $70.

I guess it falls into two camps:

1) New miners who jumped in this month with all the Bitcoin press who don't realize how fast the difficulty is going up

2) Established miners who already have everything all set up so there isn't much point of shutting down unless they are actually losing money.

Either way the good ole days of pretty profitable mining are gone (not that I was a part of this myself).  Getting into ASICS will require an investment, and they can't be resold or else-wise used like GPUs can.

 :'(  "back in the good ole days..."

I hear this comment a lot, but I don't understand it.  Given current BTC prices, the RoI on my GPUs has been higher in April '13 than any months I've been mining.  Granted I'm not an old-timer, but have been mining since Jan 2012.

GPU mining is a great way to repurpose old hardware especially if you're in an area with low electricity costs.

However the GPU mining applications don't need to go away even if costs of electricity skyrocket.

Most GPU miners use OpenCL.  There are now FPGAs on the market that can be programmed directly using OpenCL and exhibit performance exceeding that of my FPGA unit from Butterfly Labs.  More importantly, it's half the cost and uses 15 watts at full throttle.  It'll still take me a year or two to make back my initial investment.  But it is just re-using the GPU mining software.

Those GPUs will find a happy home in the altcoin mining arena, of which there are several new ones to play with.

As long as there is profits and high ROI, ASICs will try to invade these alt coins. But some alt cryptocurrencies can prevent this invasion by using different system in which ASICs not work or work slower.

The scrypt based coins are ASIC resistant (if not ASIC proof).

GPU mining is just hard to jump into now. I've hit a wall with money and I can't expand my farm to anything worthwhile yet.

As far as ASIC's go, they're developed for Bitcoin hashing, meaning that for anything else they're almost worthless which is nice if people need something to fall back on.

I wrote on some other thread earlier today that I wish ASICs never appeared. I don't mine and won't (unless I could buy and get asic in a week or two, lol) so it does not bother me too much but asics are making bitcoin community to be all that bitcoin community supposedly hate in "real word" one - big fish to become bigger and other to eat a dust from their tables. Greed, greed, greed... greed is everywhere :)

Won't even start about scam mark they brought to bitcoin.

Yep... Mass migration to Litecoin before the year is out...

*looks at math did for weekly power estimates for a radeon 5830+system (no more than 250W total) (approx. 1 target period) at current Ontario power rates (comes to 0.0761CAD per KWH after making a weighted average, then accounting for holidays that have a lower power rate. The power ontario website says to use 0.08CAD for averaging but they are lazy with their math.)*

3.20CAD

*looks at income for that one week at current exchange rate at current difficulty level (321MH/s@1GHz)*

rounded down:
0.1259 = 33.09CAD (mtgox)


*calculates % I get to keep as profit*

90%


uhhhh what?


Note: I haven't done much math like that in a while because I mine regardless because I keep the BTC and don't convert it to CAD. Electricity is just the cost of creating BTC that I absorb and I choose how much total I want to absorb monthly, regardless of how much BTC I produce.

You won't believe how fucking lucky you are.


After the great idea to trash all our nuclear plants not build new, more efficient ones, here in Germany you pay somewhere around 0.25  for a kwh.

OMG!

Here in Colorado we burn coal from the mountains and we let the smoke travel across the plains into kansas, so it's their problem.  (Isn't that where BFL is?).  Our electric company Xcel sells commercial power for $0.004 to $0.006 / KHh.  (Most people do not know that).  In order to buy electricity that cheap you need to consume 25,000 watts constantly and pay a fee of $300 US per month extra.  It is what big factories pay.  (And large marijuana farms).

If GPU miners all migrate to Litecoin then Litecoin will gain an even larger backing by the miners.

Essentially we will have a split between Bitcoin and Litecoin as the two major cryptocurrencies.

ASIC miners on bitcoin will make as much as the average GPU miner on litecoin, which practically renders ASICs as an equal to GPU. This means more miners will be likely to flock to Litecoin because building custom rigs that one can call their own is a more enjoyable hobby.

GPUS will move to litecoin. Litecoin will become honestly valuable. Someone will build an ASIC for Litecoin, and gpu miners will have to find something else to mine.

So, you think scrypt based coins are not ASIC resistant?

Honestly, I do not know.  Come join us and let's figure it out.

https://bitcointalk.org/index.php?topic=179033

I guess you don't know how memory intensive Litecoin is. The more hashes you have, the more memory you need. A 7970/7950 alone takes up 1.5-2 GB of memory. Imagine a 60 mhash ASIC hashing away.. The memory requirement is HORRENDOUS.

Yea but memory is ULTRA cheap.  Gigs cost dollars.  Now stop trying to scare the miners away from litecoin mr early adopter.

I don't see how your project benefits anyone. In fact, you're trying to accomplish the same ends as what is happening on Bitcoin. Litecoin should not have ASICs unlike Bitcoin. Actually your project is completely hypocritical to your own statements on being "ASIC hostile".

Let's see if you can fit 120 GB of memory onto a 60 mhash ASIC. That's just "RAM" as well (DDR3), factor in the memory for the actual "cores" (GDDR5). It costs about $50 for 2 GB of GDDR5, and we'll need way more than that for an ASIC.

After that, then you would have to have an ENORMOUS and super fast memory bus; otherwise it's no more efficient than a GPU.

If this is all said and done, it would be slightly more efficient compared to GPU. (2-10x more), compared to bitcoin ASICs which is a huge step over the GPUs.

if what you say is true then every miner should move to litecoin, no?

Yes, the average individual is better off mining litecoin compared to bitcoin which will become dominated by the big players. Huge quantities of ASICs may be produced by private entities for their own personal gain.

This is potentially a problem since it could cause problems such as the widespread centralization of Bitcoin mining. Open market supply for ASICs may be minimal compared to the production for personal mining.

So far so good... BUT when all that miners go to LTC.. what happens to:
LTC diff. and more importantly - PRICE, who on earth will want to buy LTC with precious little bitcoins

We all seen that when end of 2012 when block reward went half for bitcoin. IMO Litecoin is a pure speculation (4) now

Starting to mine litecoin woud be a good idea, I think. Had I not stopped mining bitcoins a while ago, I would have hit that bubble and I wouldn't be regretting falling out. An early start on any crypto is a good idea. Large investments may not be.

I still think people should invest some time into trying out bitcoin whether it's for fun or profit.

GPU mining will be completely dead within 3 months. For bitcoin at least. The alt currencies might be worthwhile to mine for but I predict their value is going to stay very low.

GPU mining will die because ASICs are an order of magnitude faster and use less power too. GPU will consume more electricity that they could ever hope to make back mining. We will see difficulties I'm the hundreds of millions. Do the math and see that GPU mining is doomed.

Math doesn't tell me the future price of bitcoin. I find what you are saying to be likely but I disagree that it is a definite possibility. For example, using Ontario electric rates, bitcoins would have to be $600-700CAD+ each to make a profit mining using most GPU setups. I personally find this unlikely but with Bitcoin, you never know. This assumes an 800TH/s network that I expect optimistically by Nov/Dec.

Did you know that at the current difficulty, CPU mining can be profitable at $290CAD/BTC at Ontario electricity prices? (total system 200W load 1090t doing 23MH/s)

Also, had people mined at the peak of 2011 via CPU and kept the coin, they could sell them now and make 4 times the cost of the electricity in profit. This assumes current-day Ontario electricity rates.

It may die for bitcoin, but miners will just keep moving on to the next profitable coin, if there are any.

We wouldn't need that, but a 1-5mh/s ASIC would be a good thing, if it consumed less power. I believe that reducing worldwide power consumption is always a good thing. And a 5mh/s LTC ASIC with around 8gigs of GDDR5 and 16gb of DDR3 would be doable.

This kind of machine wouldn't disrupt the market too much, would be quite expensive and save tons of power.